|Publication number||US6725404 B1|
|Application number||US 09/580,782|
|Publication date||Apr 20, 2004|
|Filing date||May 26, 2000|
|Priority date||May 26, 2000|
|Publication number||09580782, 580782, US 6725404 B1, US 6725404B1, US-B1-6725404, US6725404 B1, US6725404B1|
|Inventors||Apurba Choudhury, Thomas Wen Liang|
|Original Assignee||International Business Machines Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (8), Referenced by (14), Classifications (6), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The invention is generally related to functionally testing an electrical connector, and in particular, to testing a digital data bus electrical connector for reliability.
Many systems and networks of systems rely upon the communication of digital data along electrical transmission lines. The digital data has a predetermined data rate during which an electrical signal composed of pulses selectively varies between high and low voltage levels. These transmission lines interconnect all levels of electronics from integrated circuit, module, system, and network. Even systems that convey digital data in a radiant form of energy such as visible light in a fiber or radio frequency (RF) broadcast still depend in part on electrical transmission lines for a portion of the communication.
Typically, an electrical signal has clearly defined rising and falling edges that must transition within a predetermined amount of time for correct reception. Often, the electrical signal is clocked such that a data bit is transmitted at each transition of a clock signal. Failure of the electrical signal to transition from a high voltage level to a low voltage, or the reverse, prior to an associated clock signal transition results in the incorrect digital state (e.g., 1 rather than 0) being received.
The ability of an electrical transmission line to transmit data in a timely manner is often affected by the impedance in the line. Impedance is often considered to incorporate a resistive component and a reactive component, and can result in distortions in an electrical signal that delay signal transitions and potentially create data transmission errors.
Increasingly, higher data rates are used in electrical transmission lines. As such, the time frame that transitions in a signal must be received continues to decrease, making impedance an ever-increasing concern for digital transmission systems. Consequently, distortion of the electrical signal may impose an upper limit on the data rate that may be achieved within a given transmission line.
Certain types of electrical components within an electrical transmission line can affect impedance levels and line reliability. For example, faulty electrical connectors, or interconnects, can often introduce unacceptable impedance levels into an electrical transmission line. The physical mechanism for such unacceptable impedance in an electrical connector may be a reduced conducting surface area between electrical contacts due to variations in manufacturing, contamination, and material wear. With the reduced conducting surface area, the resistive nature of the electrical connector increases. This increase in resistance aggravates any reactive characteristics of the electrical connector. In addition, the reduced coupled surface area may also increase the capacitance of the electrical connector, given the increase in closely spaced, uncoupled surface areas surrounding the coupled surface area.
Because of these variations in electrical connectors, there exists a need for ensuring that an electrical connector will perform reliably in various environments. Reliability testing is often necessary since the variations may be difficult to overcome by other means, such as by choice of electrical connector design. Moreover, an electrical connector design may require validation in a different environment.
It is generally known to estimate reliability of an electrical connector by “glitch detecting”, in which a DC voltage is used to measure changes in resistance over time. If an electrical connector has a resistance value that varies more than 10 milli-Ohms (mΩ), then a failure is deemed to have occurred, based on a belief that a 10 mΩ variation means that 90% of the contact area has been lost. Reliability testing with a glitch detector is a coarse pass/fail test and not a direct indication of the suitability of an electrical connector to a specific application.
Even assuming that measuring variation in resistance of an electrical connector indicates reliability, generally known glitch detectors are subject to a number of sources of inaccuracy in measuring resistance. As the data rates required of electrical connectors increases, glitch detectors have to measure minute variations in resistance that last for correspondingly shorter periods of time. At these short durations, the resistance measurements are increasingly subject to electromagnetic interference (EMI), and thus cumbersome EMI shielding techniques must be employed.
Consequently, a significant need exists for a testing technique of an electrical connector that indicates whether a desired data rate may be reliably transmitted by the connector.
The invention addresses these and other problems associated with the prior art by providing a test apparatus and method of determining the reliability and/or suitability of an electrical connector for use at a desired data rate by sensing a propagation delay imposed by the electrical connector on a test signal.
In one aspect consistent with the invention, a test apparatus includes a connector interface and a test circuit. The test circuit generates a test signal that is transmitted through an electrical connector via the connector interface. The test circuit determines the reliability of the electrical circuit by determining a propagation delay imposed on the test signal by the electrical connector.
In another aspect consistent with the invention, a method of testing an electrical connector for reliability includes interfacing a test signal path to through the electrical connector, transmitting a test signal along the test signal path through the electrical connector, and detecting a propagation delay in the test signal after it has been transmitted through the electrical connector.
These and other advantages and features, which characterize the invention, are set forth in the claims annexed hereto and forming a further part hereof. However, for a better understanding of the invention, and of the advantages and objectives attained through its use, reference should be made to the Drawings, and to the accompanying descriptive matter, in which there is described exemplary embodiments of the invention.
FIG. 1 is block diagram of a test apparatus consistent with the invention, shown interfaced with an electrical connector under test.
FIG. 2 is a schematic diagram of one implementation of the test apparatus of FIG. 1.
FIG. 3 is an illustrative plot of a transition in a reference signal compared to two transmitted test signals, one of which is uncorrupted and another one which is corrupted by an excessive propagation delay.
FIG. 4 is a perspective view of an exemplary test box of the test apparatus of FIG. 1.
Turning to the Drawings, wherein like numbers denote like parts throughout the several views, FIG. 1 illustrates a test apparatus 10 that includes a test circuit arrangement 12 and a connector interface 14. The test circuit arrangement 12 has a plurality of test signal paths 16 and a reference signal path 18. The plurality of test signal paths 16 correspond to a plurality of electrical connector conduction paths 20 that are to be tested, depicted as being part of an electrical connector 22 functioning as a device under test (DUT).
Each conduction path 20 begins respectively with a first connector contact 24 and terminates respectively at a second connector contact 26. The connector interface 14 couples to the electrical connector 22 respectively via first test contacts 28 for the first connector contacts and second test contacts 30 for the second connector contacts.
A test signal generator 32 transmits a test signal ST through the test signal paths 16. A reference signal generator 34, synchronized to the test signal generator 32, transmits a reference signal SR through the reference signal path 18. Each test signal path 16 terminates at a delay detector 36 that compares a transmitted test signal ST′ that has been transmitted through the electrical connector 22 with the reference signal SR to detect a propagation delay.
It should be appreciated by those having skill in the art having the benefit of the instant disclosure that for clarity the number of test signal paths 16 illustrated in FIG. 1 corresponds directly to the number of electrical connector conduction paths 20. However, it is consistent with aspects of the invention for the number of number of test signal paths 16 not to correspond to the number of electrical connector conduction paths 20. In these instances, excess test signal paths 16 may be ignored or excess electrical connector conduction paths 20 tested sequentially rather than in parallel.
Moreover, the connector interface 14 may include a multiplexing circuit whereby one test signal ST from the test signal generator 32 is sequentially sent through each conduction path 20. Furthermore, the connector interface 14 may include a demultiplexing circuit that allows each of the conduction paths 20 to be sequentially delay detected by one delay detector 36.
Furthermore, although a single reference signal and a single test signal is illustrated for clarity, a plurality of reference signals and test signals may be used. For example, different conduction paths 20 within the electrical connector may be tested for a different acceptable propagation delay. As another example, the plurality of test signals and reference signals may be sequentially multiplexed for each conduction paths 20. The plurality of test signals and reference signals may be selected to represent a various operating data rates, power levels, etc., to further characterize the electrical connector 22.
In addition, the electrical connector 22 may represent a combination of components including transmission lines and additional electrical connectors.
Referring to FIG. 2, one illustrative implementation of the test apparatus 10 of FIG. 1 is depicted for reliability testing the electrical connector 22 intended for a digital data transmission line. A single test signal path 16 is shown for clarity along with a reference signal path 18, although a plurality of test signal paths 16 and corresponding delay detectors 36 may be included.
In addition, an instrumentation signal path 38 for generating an instrumentation signal SI parallels the other paths 16, 18 for tuning the test circuit apparatus 12. The instrumentation signal path 38 may be used to empirically determine the suitability of an electrical connector 22 for a specific application without necessarily having to determine what electrical characteristics of the connector 22 are required. “End-to-end testing” an electrical connector 22 achieved by delay testing may more accurately reflect reliability and avoid the sources of inaccuracy inherent in conventional testing techniques.
A pulse generator 40 provides a pulsed signal SP as a common excitation for the test signal ST the test signal path 16, the reference signal SR in the reference signal path 18, and the instrumentation signal SI in the instrumentation signal path 38. Thus, the paths 16, 18, 38 are referenced to one another. The pulse generator 40 in combination with the portions of the test signal path 16 described below thus form the test signal generator 32. Similarly, the pulse generator 40 in combination with portions of the reference signal path 18 described below form the reference signal generator 34.
The test circuit arrangement 12 advantageously includes a data enable circuit arrangement 42 that interrupts the pulsed signal SP, and thus the test signal ST, in the test signal path 16. In the illustrative test circuit arrangement 12, the data enable circuit arrangement 42 does not interrupt the reference signal path 18 and instrumentation signal path 38 so that the test circuit arrangement 12 may be tuned with the electrical connector 22 electrically isolated. In some instances, the power levels required to adequately test the electrical connector 22 may be high enough to cause safety consideration or to alter test results over time due to heating. Consequently, the data enable circuit arrangement 42 allows partial activation of the test circuit arrangement 12 for test preparation without necessarily transmitting through the electrical connector 22.
The data enable test circuit arrangement 42 includes an AND logic gate G1 in the test signal path 16. The AND logic gate G1 receives the pulsed signal SP at a first input and a data signal SD at a second input. The data signal SD goes to a high logic when DATA ON is desired by closing a data switch 44 that couples the second input to a power supply. A pull-down resistor R1 is coupled between the second input and ground.
In order to preserve the synchronization of the pulsed signal SP in each path 16, 18, and 38, the reference signal path 18 includes an AND logic gate G2 that receives the pulsed signal SP at both of its inputs. Also, the instrumentation signal path 38 has an AND logic gate G3 that receives the pulsed signal SP at both of its inputs.
The test circuit arrangement 12 includes low pass filtering of the pulsed signal SP in both the test signal path 16 and reference signal path 18. The low pass filtering serves to reduce noise as well as to introduce a predetermined delay into the reference signal SR relative to the test signal ST. A test signal filter 46 filters a data enabled test signal from AND logic gate G1 in the test signal path 16. The test signal filter 46 is formed from a series variable resistor VR1 terminated with a capacitor C1 that is coupled to ground. A first reference signal filter 48 filters a signal from the AND logic gate G2 in the reference signal path 18. The first reference signal filter 48 is formed from a series variable resistor VR2 terminated with a capacitor C2 that is coupled to ground.
In addition to delaying the reference signal SR relative to the test signal ST, the reference signal SR is also inverted relative to the test signal ST. This allows an illustrative delay detector 36 described below to respond to the failure of the test signal ST′ to complete a transition by the time that the reference signal SR completes its transition after the known delay.
The test signal path 16 thus includes an inverter 50 formed from a NOR logic gate G4 that receives at one input a signal from the first test signal filter 46. The NOR logic gate G4 has a second input coupled to the power supply.
In order to maintain the timing relationship between the paths 16, 18, a buffer 52 is formed from a NOR logic gate G5 in the reference signal path 18 receives a signal from the reference signal filter 48 at a first input. The NOR logic gate G5 has a second input coupled to ground. The buffer 52 provides a delay that corresponds to a delay by the inverter 50.
The output of the buffer 52 thereafter receives additional filtering and delay by a second reference signal filter 54 in the reference signal path 18. The filter 54 is formed from a series variable resistor VR3 terminated in a capacitor C3 coupled to ground.
A first transmission line emulator 56 in the test signal path 16 further conditions the test signal ST prior to transmission through the electrical connector 22. The emulators 56, 58 may advantageously convert voltage and current levels to be representative of the application. In addition, the transmission line emulator 56 may assist in end-to-end testing by introducing the fixed delay expected in an application. For example, the emulator 56 is formed from a series of a bus driver B1 coupled to a series resistor R3.
Similarly, the reference signal path 18 includes a second transmission line emulator 58 formed from a series of a bus driver B2 coupled to a series resistor R4. In some instances, either or both of the transmission line emulators 56, 58 may include a predetermined length of transmission line to introduce a fixed delay into either path 16, 18.
The test signal ST then transmits through the electrical connector 22 to be output as a transmitted test signal ST′. The electrical connector 22 is depicted in FIG. 2 as comprising a time-varying, passive first-order RLC (i.e., Resistor-Inductor-Capacitor) circuit of a series inductor IC(t) and resistor RC(t), between which a capacitor CC(t) is coupled to ground. The pulsed nature of the test signal ST thus interacts with the impedance of the electrical connector performing an AC dynamic test, rather than just the resistive portion of the impedance as would a DC “glitch detection” test signal.
Tuning the test circuit arrangement 12 is illustrated by an oscilloscope plot 60 triggered by the instrumentation signal SI of the instrumentation signal path that is plotting the transmitted test signal ST′ on one channel versus the reference signal SR on another channel. The oscilloscope 60 may be used when adjusting the variable resistors VR1, VR2, VR3 in preparation for reliability testing. For example, a known good electrical connector 22 may be installed, or a low impedance bypass of the electrical connector 22 be employed. The tuning of the variable resistors VR1, VR2, VR3 may then be accomplished to achieve a predetermined delay defined as the acceptable range, for instance.
The transmitted test signal ST′ is input to the delay detector 36 for comparison with the reference signal SR. In particular, a D flip flop 62 acts as a comparator by receiving the transmitted test signal ST′ at a Data (D) input and the reference signal SR at a Clock (CLK) input. The output (Q) of the D flip flop 62 is an output signal SO that is set to the logic state of the Data (D) input at each rising edge of the Clock (CLK) input. For instance, the output signal SO remains logic=0 until the transmitted test signal ST′ is still logic=1 when the reference signal SR triggers the D flip flop 62 with a rising edge transition to logic=1, in which case the output signal SO will be a logic=1 fault signal. Monitoring the output signal SO provides indications of intermittent failures since each subsequent cycle of the reference signal SR clears any fault signal from the output signal SO.
In order to latch a failure signal for human monitoring or other purposes, the delay detector 36 has an indicator latch function. More particularly, a D flip flop 64 has a Data (D) input coupled to the power supply and a clock input (CLK) coupled to the output signal SO from the comparator D flip flop 62. Thus, a latched output signal SO′ from the output (Q) of the indicator latch D flip flop 64 goes to logic=1 when a fault signal (logic=1) is present in the output signal SO. The latched output signal SO′ is current limited by a series resistor R5 and then biases a light emitting diode (LED) 66. A normally-open clear switch 68 is closed, coupling CLEAR inputs for both D flip flops 62, 64 to clear a latched indication of failure signal.
It should be appreciated by those skilled in the art having the benefit of the instant disclosure that each of the elements describe, and the order of the elements, may be omitted, or substituted with various known replacements. For example, other filtering circuit arrangements would also be appropriate, including active components and higher order filters. As a further example, the reference signal SR may be inverted rather than the test signal ST. Also, no inversion may be required if using a delay detector that triggers on the signals having opposite logic states, one high and one low, for instance.
Referring to FIG. 3, the oscilloscope plot 60 illustrates the operation of the delay detector 36. At a time t=0, a trace for an uncorrupted test signal ST′ is stable at a logic=1 state (i.e., greater than a logic=1 threshold), and a trace for a reference signal SR is stable at logic=0 state (i.e., less than a logic=0 threshold). At time t=1, the uncorrupted test signal ST′ begins a transition from a logic=1 state to a logic=0 state. At time t=2, the uncorrupted test signal ST′ has completed the transition to the logic=0 state.
The delay in the reference signal SR results in its transition continuing after time t=2, reaching logic=1 state at time t=3, at which point the state of the Data (D) is output (Q) of the comparator D flip flop 62. Since the uncorrupted test signal ST′ is logic=0 at this point, the output signal SO does not indicate a failure. A reference delay dR between time=2 and time=3 indicates a time margin available for a slower transition in the test signal ST′.
Depending on the application, the logic thresholds and upper and lower voltages of the test signal ST′ and reference signal SR may vary. Consequently, the exact delay or low pass filtering necessary to achieve this reference delay dR would vary.
Also, a trace of a corrupted test signal ST′ similarly begins a transition at time t=1, but has remained in a logic=1 state at time t=3 when the reference signal transitions to a logic=1 state. Consequently, the logic=1 state of the corrupted test signal ST′ becomes a failure signal in the output signal SO.
Referring to FIG. 4, a test apparatus 10 illustrates an environment in which a user may test an electrical connector 22, such as DIMM socket. The test circuit arrangement 12 is enclosed within a housing 70. A power switch 72 enables power to reach the test circuit arrangement 12. A plurality of knobs 74 allows adjusting delay and filtering. A ribbon cable 76 is depicted as being part of the connector interface 14, which may be of a length for convenient installation of the electrical connector 22 and may further be a selectable feature for a desired transmission line length for a fixed delay. A plurality of plugs 78 for coupling to internal signal from the test circuit arrangement 12 allow displaying the internal signals on an oscilloscope 60 or other logic tester.
It should be appreciated by those skilled in the art having the benefit of the instant disclosure that it would be possible to provide additional features to the test apparatus 10.
For example, rather than relying upon a user to monitor LEDs 66, a data acquisition system may store, analyze and alert the user of an electrical connector found to be faulty or otherwise inadequate for an intended use.
As another example, the tuning of the known delay may be unnecessary, or provided by a selection of preset values that do not require display like an oscilloscope.
As an additional example, the test circuit arrangement 12 may be incorporated into an operational electrical connector 22 and transmission line arrangement so that periodic testing may be performed without removing the electrical connector 22 from its operational installation.
As yet a further example, although a digital electrical connector 22 is included in an illustrative version described herein, the electrical connector 22 may be used in an analog implementation, such as a base band audio electrical connector 22.
As another example, in some applications the electrical connector 22 may advantageously include an impedance that generates a desired attenuation or propagation delay. The test circuit may then include detection of a propagation delay that is too short in addition to, or as an alternative to, detection of a propagation delay that is too long.
Other modifications will be apparent to one of ordinary skill in the art. Therefore, the invention lies in the claims hereinafter appended.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4893072 *||Jun 22, 1988||Jan 9, 1990||Hitachi, Ltd.||Apparatus for testing an integrated circuit device|
|US5448661 *||Apr 20, 1993||Sep 5, 1995||Hitachi, Ltd.||Optical parallel transmission device|
|US5568576 *||Jul 27, 1995||Oct 22, 1996||Hitachi, Ltd.||Optical parallel transmission device|
|US5715339 *||Sep 4, 1996||Feb 3, 1998||Hitachi, Ltd.||Optical parallel transmission device|
|US5867030 *||Jul 20, 1995||Feb 2, 1999||Advantest Corp.||Transmission path structure for measuring propagation delay time thereof|
|US6369601 *||Oct 7, 1999||Apr 9, 2002||Advantest Corporation||Method of measuring a propagation delay time through a transmission path in a semiconductor integrated circuit testing apparatus and semiconductor integrated circuit testing apparatus using the same|
|US6415402 *||Jan 24, 2001||Jul 2, 2002||International Business Machines Corporation||Programmable timing circuit for testing the cycle time of functional circuits on an integrated circuit chip|
|US6433558 *||May 11, 2000||Aug 13, 2002||Microtest, Inc.||Method for diagnosing performance problems in cabling|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US6901344 *||Feb 11, 2003||May 31, 2005||Hewlett-Packard Development Company, L.P.||Apparatus and method for verification of system interconnect upon hot-plugging of electronic field replaceable units|
|US7002355 *||Jan 8, 2004||Feb 21, 2006||Panduit Corp.||Apparatus and method for communications testing|
|US7071705 *||Dec 1, 2005||Jul 4, 2006||Panduit Corp.||Apparatus and method for communications testing|
|US7154280 *||Jun 13, 2006||Dec 26, 2006||Panduit Corp.||Apparatus and method for communications testing|
|US7395480 *||Apr 6, 2006||Jul 1, 2008||Advantest Corporation||Test apparatus and test method|
|US8093921 *||Jan 10, 2012||Cisco Technology, Inc.||Monitoring of interconnect reliability using a programmable device|
|US20040140813 *||Jan 8, 2004||Jul 22, 2004||Masud Bolouri-Saransar||Apparatus and method for communications testing|
|US20040158424 *||Feb 11, 2003||Aug 12, 2004||Mantey Paul John||Apparatus and method for verification of system interconnect upon hot-plugging of electronic field replaceable units|
|US20060082377 *||Dec 1, 2005||Apr 20, 2006||Masud Bolouri-Saransar||Apparatus and method for communications testing|
|US20060217001 *||Jun 13, 2006||Sep 28, 2006||Panduit Corp.||Apparatus and method for communications testing|
|US20060248416 *||Apr 6, 2006||Nov 2, 2006||Advantest Corporation||Test apparatus and test method|
|US20090090908 *||Oct 5, 2007||Apr 9, 2009||International Business Machines Corporation||Providing A Duplicate Test Signal Of An Output Signal Under Test In An Integrated Circuit|
|US20090091345 *||May 1, 2008||Apr 9, 2009||Moises Cases||Structure for providing a duplicate test signal of an output signal under test in an integrated circuit|
|US20100207649 *||Aug 19, 2010||Nandakumar Krishnan||In situ and real time monitoring of interconnect reliability using a programmable device|
|U.S. Classification||714/724, 714/715|
|International Classification||G01R31/04, G01R31/28|
|May 26, 2000||AS||Assignment|
Owner name: INTERNATIONAL BUSINESS MACHINES CORPORATION, NEW Y
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:CHOUDHURY, APURBA;LIANG, THOMAS WEN;REEL/FRAME:010856/0257
Effective date: 20000524
|Oct 29, 2007||REMI||Maintenance fee reminder mailed|
|Apr 20, 2008||LAPS||Lapse for failure to pay maintenance fees|
|Jun 10, 2008||FP||Expired due to failure to pay maintenance fee|
Effective date: 20080420